Decadal Observations of Internal Wave Energy, Shear, and Mixing in the Western Arctic Ocean

Abstract

As Arctic sea ice declines, wind energy has increasing access to the upper ocean, with potential consequences for ocean mixing, stratification, and turbulent heat fluxes. Here, we investigate the relationships between internal wave energy, turbulent dissipation, and ice concentration and draft using mooring data collected in the Beaufort Sea during 2003–2018. We focus on the 50–300 m depth range, using velocity and CTD records to estimate near‐inertial shear and energy, a finescale parameterization to infer turbulent dissipation rates, and ice draft observations to characterize the ice cover. All quantities varied widely on monthly and interannual timescales. Seasonally, near‐inertial energy increased when ice concentration and ice draft were low, but shear and dissipation did not. We show that this apparent contradiction occurred due to the vertical scales of internal wave energy, with open water associated with larger vertical scales. These larger vertical scale motions are associated with less shear, and tend to result in less dissipation. This relationship led to a seasonality in the correlation between shear and energy. This correlation was largest in the spring beneath full ice cover and smallest in the summer and fall when the ice had deteriorated. When considering interannually averaged properties, the year‐to‐year variability and the short ice‐free season currently obscure any potential trend. Implications for the future seasonal and interannual evolution of the Arctic Ocean and sea ice cover are discussed.

Density Staircases Are Disappearing in the Canada Basin of the Arctic Ocean

“…Finally, the recent smoothing of the staircases could potentially result from an increase of the turbulent mixing, arising as a possible consequence of the sea ice decline (Dosser et al., 2021). Yet, based on the analysis of mooring A data, Fine and Cole (2022) found that, over the past 15 years, in response to the sea ice decline, the estimated turbulent mixing between 50 and 300 m did not increase, despite an increase of the near‐inertial energy. This is because waves are forced with large vertical wavelength, and consequently have low vertical shear.…”

Winter Vertical Diffusion Rates in the Arctic Ocean, Estimated From 7Be Measurements and Dissipation Rate Profiles

“…The combined effect of receding ice cover and other trends in the Arctic System, for example, changes in the freshwater dynamics, on turbulent transport is not conclusively understood, and might be regionally and seasonally different. No pronounced trends are reported in the western Arctic Ocean (Dosser et al., 2021; Fine & Cole, 2022), while stronger changes are observed in the eastern Arctic Ocean, especially during summer season (Dosser et al., 2021). The region along the continental slopes of the eastern Arctic Ocean, from the Kara Sea to the Siberian Seas, is subject to Atlantification (Polyakov et al., 2017), that is, an eastward progression of conditions typical for the Atlantic Ocean.…”

Validating Finescale Parameterizations for the Eastern Arctic Ocean Internal Wave Field

“…The outlook of the application of FS in a changing Arctic Ocean is complex. Although the declining sea-ice cover is associated with an increase in near-inertial energy input, this does not currently translate to an observable increase in internal-wave driven turbulence (Fine & Cole, 2022). However, the ongoing "Atlantification" of the eastern Arctic Ocean also leads to a decrease in stratification (Polyakov et al, 2017(Polyakov et al, , 2018.…”

“…Applications of this formulation include global estimates of dissipation rates from the ARGO buoy fleet (Whalen et al., 2015), ice‐tethered profilers in the central Arctic (Dosser et al., 2021), and CTD profiles in Canadian Arctic Shelf waters (Chanona et al., 2018). Fine and Cole (2022) point to a significant caveat of this approach: While shear variance can explain 55% of their observed variability of R ω , strain variance only accounts for 10%, making the FS predictions highly dependent on the choice of R ω . In this study, we calculate the “strain‐only” version of FS to validate it against the “full” (shear & strain) FS and direct observations of dissipation rate.…”